Six studies (n=221) were included in the analysis of endoscopic ultrasound-guided coeliac plexus block for chronic pancreatitis. There were four prospective studies (three were randomised, one of these was blinded) and two retrospective studies. Average follow-up ranged from four to eight weeks (stated in three studies). Three studies (n=119) were included in the analysis of endoscopic ultrasound-guided coeliac plexus neurolysis for pancreatic cancer pain. There were two prospective studies and one retrospective study. Follow-up periods appeared to range from two weeks to 52 weeks (there was some discrepancy in the reporting).
The pooled estimated proportion of patients with pain relief for chronic pancreatitis from endoscopic ultrasound-guided coeliac plexus block was 51.46% (six studies). The 95% CI suggested that the true proportion who experienced pain relief was over 32.72%. The authors suggested that the procedure had some efficacy in these patients.
The pooled estimated proportion of patients with pancreatic cancer pain relief from endoscopic ultrasound-guided coeliac plexus neurolysis was 72.54% (three studies). The 95% CI suggested that the true proportion of patients who benefited from the procedure was over 50.62%. The authors suggested that this represented good efficacy.
The most frequently reported adverse effects for endoscopic ultrasound-guided coeliac plexus block were transient diarrhoea, transient orthostatic hypotension, transient increases in pain and abscess formation. For endoscopic ultrasound-guided coeliac plexus neurolysis, the most common adverse effect was transient hypotension; there were no major neurologic complications.
There was no evidence of statistical heterogeneity. Publication bias was not indicated.